So a few weeks ago I was searching for a focus for my dissertation and after a conversation with Margaret Sambell, head of strategy at e-skills UK, I’ve identified a subject which really intrigues me:

Why do Computer Science graduates consistently have the highest rates of unemployment by subject (as reported in the HESA Destinations of Leavers (DLHE) survey), despite widespread skills shortages within the IT Industry?

Every year the HESA DLHE stats are released and the press reports the facts – Computer Scientists have the highest unemployment rates. However alongside this headline figure, there’s usually a lot of speculation about the reasons for this figure, very little of which is based upon fact. Some years the students are blamed – they have poor soft skills and can’t operate within the real world (e.g. headlines focusing on “geek speak”). Other times it is the universities that are to blame – the courses don’t teach the skills wanted by employers. In the media coverage, rarely has there been an attempt to truly understand the problem and the main factors behind the headline figure.

The Council of Professors and Heads of Computing (CPHC) have attempted to understand the issue, commissioning a report conducting an in-depth analysis of the HESA DLHE data, which looked for some of the factors hidden within the figures. Their findings suggest that specific demographic factors of the computer science cohort, such as university type (e.g. Russell Group, verses other pre-1992, verses post-1992 universities), ethnicity and degree class may all contribute to the high unemployment rates. The percentage of students studying at post-1992 universities is much higher for computer science than other subjects and the proportion of BME students is also higher than the student population as a whole. As students from both of these groups have higher rates of unemployment, it makes sense that the overall picture for computer science students is worse. Another factor considered was degree class and the smaller numbers of students achieving a “good degree” (2.1/1st) within computer science could mean that more students end up unemployed.

The thing I found fascinating though that even accounting for these factors, there are still wide variations in employment rates across different universities within the same “group”. One example given in the report was from the North East, where Teeside (29% unemployed), Sunderland (12%) and Northumbria (8%) – all post-1992 universities, with a similar ethnicity profile (mainly white), but considerable differences in unemployment rates. If asked to predict the rates for these three universities I would have got the order and magnitude right. Instinctively I think of Northumbria as being a “better” university than Teeside. In this case, it may simply be down to the economic fortunes of Middlesbrough, compared with Newcastle, but I wonder to what extent institutional bias is also a factor. I wonder whether someone’s sense of what makes a good university comes out of the DLHE statistics and employment rates? If so, are the unemployment rates driven by the assumptions we have about the institution from which someone has studied? Do employers reject someone simply because they want to Teeside, rather than Northumbria? Do some students choose to go to Northumbria, instead of Teeside because they think it will increase their chance of getting a job? If either of these are true, does that mean students who go to Teeside don’t care about their employability and are therefore less likely to get a job at the end? I wonder whether a university with a low employment rate can improve this or not. Lots of interesting questions. It’s a complex issue and it feels like a lot of the issues are self-perpetuating. It’s very easy to go around in circles and not find any useful answers.

In my research, I’ve decided to focus upon student or graduate attitudes towards their own employability. I want to understand why some students choose to take steps to improve their employability and why others don’t. Why do some students undergo work placements and others refuse them? Do students consider employment rates when choosing their university course? I want to understand whether Computer Science graduates have different attitudes to students of other subjects. I want to understand whether those attitudes vary amongst students from different universities and whether that may explain the different employment rates.

At the moment I’m still unsure of my approach. I really want to survey or interview graduates from computer science and other STEM courses who graduated in 2013 (the last population to complete the DLHE survey), but I don’t know where to find them. I am keen to do a case study of the three different Manchester-based universities, University of Manchester, Manchester Met and University of Salford as they provide representation from the three main different university groups (Russell Group, Post-1992 and pre-1992 “plate-glass” university respectively) and by using the same location I would reduce one of the variables which may influence the results, but I am worried I won’t be able to get enough participants. I may have to interview current final year students instead of graduates or I may need to open it up to more universities or year groups. We shall see – I plan to approach the universities and see whether they can advertise to their ex-graduates my research. Fingers crossed some of them may be interested in participating.

If you have any suggestions for how I can get hold of participants, please do let me know. If you are interested in this research and have any suggestions, I’m keen to hear from you.

Update – For more information about my research visit the overview page: MSc Research – Computer Science Graduate Unemployment